Yahoo plans to release some technologies, including storage technologies, to the open source community, a senior executive of the company said.
These are systems that Yahoo built to help it handle large numbers of users on its websites, but that don't necessarily give it a competitive advantage, said David Chaiken, chief architect at Yahoo, in an interview in Bangalore on Friday.
[ Get your websites up to speed with HTML5 today using the techniques in InfoWorld's HTML5 Deep Dive PDF how-to report. | Learn how to secure your Web browsers in InfoWorld's "Web Browser Security Deep Dive" PDF guide. ]
In 2009, Yahoo donated its Traffic Server scalable caching proxy to the Apache Software Foundation.
A working group on open source at Yahoo is currently evaluating technologies that could qualify to be released to open source.
The company has to first make sure that each of the technologies will really be useful and provide significant value outside Yahoo, before releasing it to open source, Chaiken said. It takes time and effort to go through the open source process, and to build a community around open source, so the company has to first make sure there will be interest from developers, he added.
Releasing technology to the open source community helps Yahoo build recognition and a technical brand in the technical community, as well as develop relationships with universities and companies, Chaiken said. There could also be some financial benefits in getting community developers to work on a project, he added.
A large part of Yahoo's IT infrastructure is built around open source software, including the Hadoop scalable distributed computing platform, which helps Yahoo execute at a very large scale, Chaiken said.
Yahoo said earlier this year that it is discontinuing its distribution of Hadoop and will instead focus on Apache Hadoop, a project of the Apache Software Foundation.
Hadoop is likely to be critical as Yahoo attempts to reduce diversity and unify its IT infrastructure. The company is for example standardizing its asynchronous or batch computation and storage using Hadoop, Chaiken said.
As a result of acquisitions and in-house projects that were rushing to bring new products and technologies to market, Yahoo over the years ended up with a lot of different systems that in some cases are fulfilling the same purpose, Chaiken said.
Each of the different science groups that were writing algorithms for extracting signal from the company's data were essentially operating on their own dedicated analytics infrastructure and storage, each building different models in a slightly different way, and each operating on different data sets, Chaiken said.
In the last four years, Yahoo has integrated the infrastructure for the science groups across the world, and removed the data and algorithm silos, he added.
Yahoo is also working on unifying its content and communications systems, on top of a "content agility platform" that depends to a large extent on Hadoop, Chaiken said. The company is in the process of unifying its media and communications properties.